Figure 1: Text and sentiment analysis for environmental problems
Course description
This course will cover foundations and applications of natural language processing. Problem sets and class projects will leverage common and emerging text-based data sources relevant to environmental problems and will build capacity and experience in common tools, including text processing and classification, semantics, and natural language parsing.
Instructor
Mateo Robbins (mjrobbins@ucsb.edu)
- Office: MEDS Group Office
- Office hours: Monday 1:15pm -2:15pm
Important links
Weekly course schedule
Lectures: NCEAS classroom
M 12:00pm -1:15pm
W 11:00am - 12:15pm
Learning objectives
The goal of EDS 231 (Text and Sentiment Analysis for Environmental Problems) is to expose students to a range of text and sentiment analysis data sources, techniques and tools that can be applied to environmental problems. During this course, students will:
Become familiar with the R packages used in text-as-data applications
Conduct and explain each step in the text data collection, analysis, and presentation pipeline
Course requirements
Computing
Minimum MEDS device requirements (bring to all sessions + charger!)
Up-to-date R and RStudio
Python version 3.x installed (although most, if not all, work will be done in RStudio)
Textbook
- Readings will be drawn from free online ebooks:
Topics
| Week | Session | Lecture/Demo | Reading | Assignment |
|---|---|---|---|---|
| 1 | 4/01 | Course Intro and Text Analysis Overview | ||
| 4/03 | NYT Lab - Key | TMR 1.0-1.3, Appendix A | Lab 1 | |
| 2 | 4/08 | Sentiment Analysis I | TMR 2.0-2.7 | In-class demo |
| 4/10 | Sentiment Analysis I Lab - key | Lab 2 | ||
| 3 | 4/15 | Topic Analysis Lecture | TMR 6.0-6.4 | |
| 4/17 | Topic Analysis Lab - key | Lab 3 | ||
| 4 | 4/22 | Classification Lecture | SMLTR 7.1-7.4 | In-class demo |
| 4/24 | Classification demo cont. | Lab 4 | ||
| 5 | 4/29 | Break | ||
| 5/01 | Break | |||
| 6 | 5/08 | Word Embeddings Lecture | SMLTR 5.1-5.7 | |
| 5/10 | Word Embeddings Lab | Lab 5 |
